Virtual Microscopy in Prostate Histopathology: Simultaneous Viewing of Biopsies Stained Sequentially With Hematoxylin and Eosin, and α-Methylacyl-Coenzyme A Racemase/p63 Immunohistochemistry
2006; Lippincott Williams & Wilkins; Volume: 175; Issue: 2 Linguagem: Inglês
10.1016/s0022-5347(05)00164-3
ISSN1527-3792
AutoresH Helin, Mikael Lundin, Mervi Laakso, Johan Lundin, Heikki J. Helin, Jorma Isola,
Tópico(s)Cell Image Analysis Techniques
ResumoYou have accessJournal of UrologyAdult urology1 Feb 2006Virtual Microscopy in Prostate Histopathology: Simultaneous Viewing of Biopsies Stained Sequentially With Hematoxylin and Eosin, and α-Methylacyl-Coenzyme A Racemase/p63 Immunohistochemistry Henrik O. Helin, Mikael E. Lundin, Mervi Laakso, Johan Lundin, Heikki J. Helin, and Jorma Isola Henrik O. HelinHenrik O. Helin Institute of Medical Technology, University of Tampere, Tampere , Mikael E. LundinMikael E. Lundin Biomedical Informatics Group, HUSLAB, Helsinki University Central Hospital, Helsinki , Mervi LaaksoMervi Laakso Department of Pathology, Seinäjoki Central Hospital, Seinäjoki, Finland , Johan LundinJohan Lundin Biomedical Informatics Group, HUSLAB, Helsinki University Central Hospital, Helsinki , Heikki J. HelinHeikki J. Helin Department of Oncology, University of Helsinki and Division of Pathology, HUSLAB, Helsinki University Central Hospital, Helsinki , and Jorma IsolaJorma Isola Institute of Medical Technology, University of Tampere, Tampere View All Author Informationhttps://doi.org/10.1016/S0022-5347(05)00164-3AboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookTwitterLinked InEmail Abstract Purpose: Histopathological diagnosis of small focus carcinomas in prostatic needle biopsies is often assisted by IHC. To make a definitive diagnosis the pathologist must compare IHC findings with hematoxylin and eosin stained tissue morphology. We introduce what is to our knowledge a new application of virtual microscopy, in which hematoxylin and eosin, and IHC stains done sequentially on the same microscope slide can be simultaneously displayed and compared on a computer screen. Materials and Methods: A total of 30 hematoxylin and eosin stained prostatic needle biopsies were scanned with a computer controlled microscope. The slides were destained and then immunostained with a cocktail of AMACR and p63 antibodies, which labels the nuclei of nonmalignant basal cells (p63) and the cytoplasm of neoplastic glandular cells suspicious for malignancy (AMACR). The slides were then scanned again and the pairs of virtual slides were aligned for synchronized viewing. Results: The presented technique was found helpful when suspicious lesions were small and when examining the immunoprofile of specimens was warranted, in addition to examining hematoxylin and eosin stained tissue morphology. The usefulness of our approach based on virtual microscopy can be evaluated on the website http://www.webmicroscope.net/AMACRp63, which also serves as an educational tool for self-learning the correlation between hematoxylin and eosin stained tissue morphology, and AMACR/p63 IHC in prostate biopsies. Conclusions: The technology for simultaneously viewing sequentially hematoxylin and eosin and IHC stained prostate biopsies can be readily used for educational purposes, as exemplified by our website, and along with the availability of rapid virtual slide scanners it can also be used for clinical diagnostics. Histopathological diagnosis of prostatic needle biopsies can be difficult due to benign mimics of cancer, eg post-atrophic hyperplasia, mucinous metaplasia and atypical adenomatous hyperplasia, PIN and borderline changes, often termed ASAP.1–3 To better discriminate these lesions from small focus carcinomas pathologists often use IHC, in addition to standard hematoxylin and eosin staining. A central histopathological criterion is basal cells, which distinguishes benign prostate glands and PIN from cancerous glands. Basal cells can be difficult to identify with a standard morphological stain but with IHC using antibodies against high molecular weight cytokeratins4-6 and the transcription factor p637–10 their presence or absence can be ascertained. The usefulness of basal cell IHC as an aid in the interpretation of prostate biopsies is well established.4–10 Since the transcription factor p63 is a nuclear protein, it can be used simultaneously with the cytoplasmic marker AMACR (clone P504S, Dako Cytomation, Copenhagen, Denmark), which was recently discovered to be up-regulated in prostate cancer.11 An increasing number of studies have described the usefulness of AMACR IHC for the differential diagnostics of prostate carcinoma.12 More recently IHC using a cocktail of AMACR and p63 antibodies has pinpointed more specifically the nature of suspicious glands in prostatic biopsies.13,14 Positive cytoplasmic AMACR reactivity occurring without nuclear p63, ie glands lacking basal cells, gives the most comprehensive immunophenotypic evidence of carcinoma.13,14 PIN is characterized by AMACR positivity in the presence of p63 positive basal cells.13,14 Although the usefulness of AMACR and basal cell markers has been firmly documented, the concordance between IHC and a diagnosis based on hematoxylin and eosin staining is not always perfect.1,2 Therefore, the histopathological diagnosis of prostate biopsies cannot be based on IHC alone. Hematoxylin and eosin staining allows the examination of cellular morphology, especially for nucleoli, which are a central cellular criterion for carcinoma. Thus, it is well established that hematoxylin and eosin, and IHC complement each other and must be evaluated in parallel. Unfortunately suspicious lesions in prostatic needle biopsies are often small and not always present on deeper sections cut from the paraffin block for IHC.15 A solution to the problem is to routinely prepare intervening unstained slides from prostate biopsy blocks. The extra slides are stored and used for IHC should the morphological stain (hematoxylin and eosin) not provide a definitive diagnosis.16 Another approach is to destain a previously stained hematoxylin and eosin slide, followed by IHC.15,17 Due to destaining hematoxylin and eosin morphology is no longer available for comparison with IHC unless a digital imaging system is used.15 We describe an application of virtual microscopy,18–20 which allows simultaneous on-screen visualization of a hematoxylin and eosin and AMACR/p63 double IHC staining done sequentially on the same slide. The entire biopsies can be viewed at any magnification in a transparent overlay mode in a single viewing window or side by side in 2 synchronized viewing windows. This method has several practical advantages, namely in histopathology education, in second opinion and quality control situations, and in clinical diagnostics when the morphology and immunophenotype of small lesions must be compared. Materials and methods For the study we selected 23 cases with formalin fixed, paraffin embedded prostatic needle biopsies available in the archive at the Pathology Department at Central Hospital Seinäjoki, Finland. Specimens were selected that favored diagnostically challenging diagnoses, such as small focus carcinoma, PIN, ASAP, proliferative inflammatory atrophy or suspicion for carcinoma without a more detailed description. All diagnoses were based on the original hematoxylin and eosin staining only. Common diagnostic entities were also included. Ordinary 3 to 4 μm tissue sections were cut from the paraffin blocks on charged SuperFrost™ Plus slides to avoid detaching tissue sections from the slides. The slides were stained routinely with hematoxylin and eosin, and digitized using a virtual microscopy slide scanning system, as described. After slide scanning the coverslips were removed by soaking the slides in xylene until the coverslips detached. The slides were then washed with absolute ethanol, air dried and immersed in antigen retrieval buffer, composed of 0.5M tris and 1 mM ethylenediaminetetraacetic acid, pH 9. Antigen retrieval was done in an autoclave at 105C for 5 minutes, followed by a 20-minute cooling period. This procedure also removes hematoxylin and eosin staining completely. After rinsing and endogenous peroxidase quenching primary antibody incubation was done using a cocktail of antibodies to p63 (clone 4A4+Y4A3, Novocastra, Newcastle, United Kingdom), dilution 1:200, and to AMACR (clone P504S), dilution 1:200, for 30 minutes at room temperature. Antibodies were detected using a PowerVision+™ reagent kit according to manufacturer instructions. The slides were immersed for 10 minutes in diaminobenzidine and enhanced with 0.5% copper sulfate for 5 minutes. The slides were counterstained in hematoxylin, dehydrated in graded ethanols, cleared in xylene and coverslipped. The now IHC stained slides were then digitized a second time. Figure 1 shows the slide staining and scanning protocol. Fig. 1. Preparation and viewing modes of virtual slide pair. H&E stained glass slide is scanned using computer controlled microscope. Same slide is destained as result of antigen retrieval procedure and immunostained with cocktail of AMACR and p63 antibodies. IHC stained slide is scanned again, and H&E and IHC virtual slides are processed for synchronized viewing layered with blending on top of each other or side by side. Virtual microscopy An Axioskop2 MOT microscope (Zeiss GmBH, Göttingen, Germany) was equipped with a NeoFluar® oil 40× objective (numerical aperture 1.3) and motorized specimen stage. A contiguous array of digital images covering the entire biopsy was captured at 0.3 μm per pixel resolution with a color sensor camera (capture resolution 1,300 × 1,030 pixels in 3 color scanning mode). The image capture process, ie stage movement, autofocus, shading correction and image capture, was automated using KS400 software, version 3.0 (Zeiss GmBH). The acquired image files were digitally sharpened and stitched into a single montage file, which was compressed into a wavelet-type image file (enhanced compressed wavelet format) using ER Mapper software (Earth Resource Mapping Pty, West Perth, Australia). The compressed virtual slides were uploaded to a web server running Image Web Server software (Earth Resource Mapping Pty, West Perth, Australia). The virtual slide pairs (hematoxylin and eosin, and IHC) were exactly aligned for synchronized viewing by keeping the hematoxylin and eosin virtual slide as a reference and adjusting the position of the corresponding IHC virtual slide (http://www.webmicroscope.net/AMACRp63/methods). Virtual slides on the website can be viewed in a standard web browser (Microsoft® Internet Explorer or Mozilla® Firefox™) on any Windows® platform. A small (600 kb) plug-in for the browser is automatically downloaded and installed at the first viewing session. Client computer requirements are modest, that is a 1 GHz standard personal computer is recommended. Results Routinely hematoxylin and eosin stained slides were first digitized using an established virtual microscopy scanning technique. The coverslips were removed and immunohistochemical staining for AMACR and p63 was performed using a standard protocol. The antigen retrieval procedure removed the hematoxylin and eosin staining completely and the quality of immunostaining on destained slides was similar to that on ordinary unstained material. Removing the coverslip and performing AMACR/P63 immunostaining requires approximately 3 hours, followed by slide scanning, which requires 0.5 to 1 hour. Figure 1 shows the slide staining and scanning protocol. A total of 30 biopsies from 23 patients were scanned and a publicly open website containing the virtual slides was created (http://www.webmicroscope.net/AMACRp63). The virtual slides representing pairs of hematoxylin and eosin, and IHC stains can be viewed within a standard web browser (Internet Explorer or Mozilla® Firefox™). Figure 2 shows a screen shot. Fig. 2. Screen shot of 2 slide virtual microscopy interface demonstrates prostatic biopsy stained with H&E (left) and AMACR/p63 immunostaining (right). Zooming (magnification change) and navigation in sample are synchronized, ie every movement occurs in H&E and IHC viewing windows simultaneously and similarly. AMACR positive and p63 negative glands suspicious for carcinoma. Inset (upper right) shows overview image of current position and extent of large window. Circle indicates ROI. By clicking on a biopsy thumbnail image on the website a new browser window is opened and the corresponding IHC, and hematoxylin and eosin virtual slides are loaded into a single viewing window but in separate layers. The viewing interface includes basic virtual microscopy functions, such as mouse and cursor navigation, and zooming to any magnification level. A small overview window is always present in the margin, showing the current view area and allowing rapid dragging navigation. The contrast and brightness of the virtual slide can be adjusted in the browser. Initially only hematoxylin and eosin staining is visible in the viewing window but hematoxylin and eosin, and IHC image blending can be controlled using a mouse operated slider bar. Almost perfect alignment of hematoxylin and eosin, and IHC staining was achieved, allowing comparison of hematoxylin and eosin morphology and immunophenotype even at a single cell level. As an alternative to blending hematoxylin and eosin, and IHC images, the user can choose to view hematoxylin and eosin, and IHC virtual slides side by side in 2 viewing windows (fig. 2). Also, in this mode zooming (magnification change) and navigation within the sample are synchronized, ie every movement takes place in the hematoxylin and eosin, and IHC viewing windows simultaneously and similarly. The website not only demonstrates the simultaneous viewing of sequentially stained virtual slides, but also serves as an educational tool for self-learning the correlation between prostate histopathology and the interpretation of AMACR/p63 immunostaining. To improve the pedagogic value an experienced pathologist (HJH) defined and annotated ROIs in slides with foci of particular diagnostic interest. ROIs are indicated on the screen by circles. They represent carcinomatous and suspicious glands, PIN, ASAP and post-atrophic hyperplasia. ROI indicators and the annotations are initially hidden but they can be turned on by the user. By clicking on an annotation hyperlink in the right margin the corresponding ROI is automatically zoomed into the viewing window and a textual description is displayed. A general comment on each slide is also available but not the definitive diagnoses of the case, which would require inspection of all biopsies. Depending on observer experience with prostate pathology the nature of most of these lesions can be diagnosed or suspected on hematoxylin and eosin staining. However, the aid of parallel AMACR/p63 immunostaining is already apparent after a short time of use. Discussion Immunostains for high molecular weight cytokeratin, p63 and AMACR are frequently used to aid in the interpretation of prostatic biopsies containing glands suspicious for malignancy. Despite the obvious advantages of IHC it is not 100% sensitive or specific to provide a definitive diagnosis of prostatic biopsies alone.1,2 Unfortunately biopsies often contain only a few suspicious glands, which may have been cut through and, therefore, are no longer present on deeper sections cut from the paraffin block for IHC.15 For this reason obtaining immunohistochemical information on such a lesion can be difficult. Our approach is based on virtual microscopy. An entire hematoxylin and eosin stained biopsy is digitized, destained and then after immunostaining the same slide is scanned a second time. Using virtual microscopy the suspicious glands identified by hematoxylin and eosin staining, and their AMACR/p63 immunophenotype can be compared on a computer screen more easily than with any other approach. To our knowledge this is the first demonstration of the use of whole slide virtual microscopy for real-time comparison of 2 stains made sequentially on the same slide. Feedback from clinical pathologists who tested the simultaneous hematoxylin and eosin, and IHC viewing was exclusively positive. If an analogous comparison is done using stored intervening unstained sections for IHC,16 the hematoxylin and eosin, and IHC slides must be changed back and forth in the microscope, and relocating the cells or gland of interest can be difficult and time-consuming. Another approach is destaining the hematoxylin and eosin stained slide to be used for IHC.15 This procedure is similar to ours but since the hematoxylin and eosin staining is removed, a real-time comparison of tissue morphology with IHC is not possible. A modification of this protocol, termed tissue protection immunohistochemistry, was recently introduced.17 In this method all except 1 tissue section on the slide are covered with a liquid cover glass embedding medium, which protects the sections from destaining. The uncovered section is immunostained and can be compared with the adjacent hematoxylin and eosin stained sections on the same slide.17 Although the procedure has proved usable, it has the disadvantage that most immunostaining robots (autostainers) cannot perform IHC on slides covered partially with a staining protectant. Furthermore, only adjacent sections and not exactly the same section on the slide can be compared. In our virtual microscopy based system the laboratory protocols need not be altered from those used routinely for IHC of unstained slides and the entire biopsy is available for real-time comparison. Since the alignment of the hematoxylin and eosin, and IHC staining is done by adjusting the on-screen positioning of the virtual slides, the presented system for synchronized viewing should be applicable to any slide scanner with a compatible output image format. Along with the availability of rapid slide scanners hematoxylin and eosin stained slides can routinely be scanned within minutes, followed by IHC staining, which typically requires 2 to 4 hours, a second virtual slide scanning, requiring minutes, and processing of the virtual slide pair for on-screen viewing on the Internet, requiring 0.5 to 1 hour. The pathologist can then view the hematoxylin and eosin, and IHC virtual slide pair on the office computer screen via the Intranet (a local computer network). Overall with recent advances in microscope slide scanning techniques and Internet based image transfer technology virtual microscopy has emerged as an important new tool, especially for histopathology education and quality control.18 The main advantage of virtual microscopy compared to conventional printed or digital photomicrographs is the possibility of viewing entire specimens, ie any part of a whole slide at any magnification. This is particularly important for the evaluation of prostatic biopsies, when assessing the architectural growth pattern of the glands is decisive for the diagnosis. Slides are first inspected at relatively low magnification but the diagnosis requires the evaluation of cellular details at high magnification. We have previously applied virtual microscopy to assess observer variation in Gleason grading.19 An excellent agreement between Gleason grades assessed from conventional glass slides and grades assessed from virtual slides indicated that virtual slides of prostatic biopsies can be used without compromising grading accuracy.19 The web based viewing technology allows the distribution of specimens to an unlimited number of pathologists for educational and quality control purposes as well as for a second opinion in problem cases. The application presented adds simultaneous viewing of 2 slides as a new functionality that cannot be done with a conventional microscope. Conclusions The presented technique for simultaneous viewing of hematoxylin and eosin staining, and immunostaining in prostate needle biopsies can readily be applied in pathology education, and in second opinion and quality control situations. With the availability of new, more rapid slide scanners the technique could also be used in clinical diagnostics. The usefulness of the virtual microscopy based approach can be evaluated on our publicly open website (http://www.webmicroscope.net/AMACRp63). In addition to prostate biopsy pathology, the technology described can be applied in any field of histopathology in which a comparison of tissue morphology and immunophenotype at high resolution is of value. References 1 : Acinar adenocarcinoma. In: WHO Classification of Tumours. Edited by . Lyon: IARC Press2004. Google Scholar 2 : Diagnosis and reporting of limited adenocarcinoma of the prostate on needle biopsy. Mod Pathol2004; 17: 307. Google Scholar 3 : Report of the Pathology Committee: false-positive and false-negative diagnoses of prostate cancer. BJU Int2003; 92: 62. Google Scholar 4 : Use of keratin 903 as an adjunct in the diagnosis of prostate carcinoma. Am J Surg Pathol1989; 13: 389. Google Scholar 5 : Usefulness of immunoperoxidase staining with high-molecular-weight cytokeratin in the differential diagnosis of small-acinar lesions of the prostate gland. Virchows Arch A Pathol Anat Histopathol1990; 417: 191. Google Scholar 6 : Utility of immunohistochemical detection of high molecular weight cytokeratin for differential diagnosis of proliferative conditions of the prostate. Int J Urol1998; 5: 237. Google Scholar 7 : p63 is a prostate basal cell marker and is required for prostate development. Am J Pathol2000; 157: 1769. Google Scholar 8 : p63 protein expression is rare in prostate adenocarcinoma: implications for cancer diagnosis and carcinogenesis. Urology2001; 58: 619. Google Scholar 9 : Comparison of the basal cell-specific markers, 34betaE12 and p63, in the diagnosis of prostate cancer. Am J Surg Pathol2002; 26: 1161. Google Scholar 10 : Diagnostic utility of immunohistochemical staining for p63, a sensitive marker of prostatic basal cells. Mod Pathol2002; 15: 1302. Google Scholar 11 : Identification of differentially expressed genes in human prostate cancer using subtraction and microarray. Cancer Res2000; 60: 1677. Google Scholar 12 : Alpha-methylacyl CoA racemase (P504S): overview and potential uses in diagnostic pathology as applied to prostate needle biopsies. J Clin Pathol2003; 56: 892. Google Scholar 13 : Use of p63/P504S monoclonal antibody cocktail in immunohistochemical staining of prostate tissue. Appl Immunohistochem Mol Morphol2004; 12: 75. Google Scholar 14 : Diagnostic utility of a p63/alpha-methyl-CoA-racemase (p504s) cocktail in atypical foci in the prostate. Mod Pathol2004; 17: 1180. Google Scholar 15 : Efficacy of restaining prostate needle biopsies with high-molecular weight cytokeratin. Hum Pathol2000; 31: 1155. Google Scholar 16 : Use of intervening unstained slides for immunohistochemical stains for high molecular weight cytokeratin on prostate needle biopsies. Am J Surg Pathol1999; 23: 567. Google Scholar 17 : Tissue protection immunohistochemistry: a useful adjunct in the interpretation of prostate biopsy specimens and other selected cases in which immunostains are needed on minute lesions. Am J Clin Pathol2002; 117: 194. Google Scholar 18 : Virtual microscopy. J Clin Pathol2004; 57: 1250. Google Scholar 19 : Web-based virtual microscopy in teaching and standardizing Gleason grading. Hum Pathol2005; 36: 381. Google Scholar 20 : digital atlas of breast histopathology: an application of web-based virtual microscopy. J Clin Pathol2004; 57: 1288. Google Scholar © 2006 by American Urological AssociationFiguresReferencesRelatedDetails Volume 175Issue 2February 2006Page: 495-499 Advertisement Copyright & Permissions© 2006 by American Urological AssociationKeywordsprostatealpha-methylacyl-CoA racemasebiopsytelepathologyprostatic neoplasmsMetricsAuthor Information Henrik O. Helin Institute of Medical Technology, University of Tampere, Tampere More articles by this author Mikael E. Lundin Biomedical Informatics Group, HUSLAB, Helsinki University Central Hospital, Helsinki More articles by this author Mervi Laakso Department of Pathology, Seinäjoki Central Hospital, Seinäjoki, Finland More articles by this author Johan Lundin Biomedical Informatics Group, HUSLAB, Helsinki University Central Hospital, Helsinki More articles by this author Heikki J. Helin Department of Oncology, University of Helsinki and Division of Pathology, HUSLAB, Helsinki University Central Hospital, Helsinki More articles by this author Jorma Isola Institute of Medical Technology, University of Tampere, Tampere More articles by this author Expand All Advertisement Advertisement PDF downloadLoading ...
Referência(s)